असतो मा सद्गमय · तमसो मा ज्योतिर्गमय · मृत्योर् मा अमृतं गमय (Lead me from the unreal to the real, from darkness to light, from death to immortality)
Bṛhadāraṇyaka Upaniṣad (1.3.28)

Intelligence that will eat the world.

Shunya Research

Who we are

We are a student-led research lab based in India, dedicated to advancing the interdisciplinary study of embodied intelligence through integrated frameworks spanning robotics, world modeling, neuroscience, and philosophy of mind.

We believe AI is fundamentally a problem about how minds work—drawing from neuroscience, cognitive psychology, and philosophy—rather than just statistical optimization. While this approach is different from mainstream AI research, we think it's the path to genuine intelligence.

Research in Progress

Embodied Intelligence & World Models

Embodied intelligence is studied through agents that learn predictive models of their environments from interaction. The emphasis is on acquiring compact world models from sensory streams such as vision, action, and proprioception, and using these models for planning and control without dense rewards or environment-specific tuning.

Experiments are conducted primarily in simulated and game-based environments, with an emphasis on understanding which learned representations support generalization and long-horizon reasoning.

Status: Prototypes and exploratory experiments

Verifiable "Thinking" in Language Models

Language models are examined through the lens of formal systems such as logic, type theory, and program semantics. This direction investigates how symbolic structure can be integrated with learned representations to produce reasoning that is internally consistent and mechanically verifiable.

The motivation is the development of language-based systems whose outputs can be checked and audited, particularly in scientific and safety-critical contexts.

Status: Prototypes and exploratory experiments

Meet the team

Sachit

Sachit Ramesha Gowda

Computer Vision and Neuroscience

Mihir

Mihir Arya

Natural Language Processing and Multimodal Reasoning

DhruV Patankar

Dhruv Patankar

Reinforcement Learning and Information Theory

Interested in Supporting Us?

Consider donating to our research initiatives:

BTC: bc1qta2vkrwp42dlrfv5qfzf34tr59k902vd7xvkl7
ETH: 0x5B04Fc2787D6e9642257BE3B8d8847075eb055E8